How important are "stocks" (steals=blocks)?

In The Book of Basketball, Bill Simmons makes a case for the greatness of Hakeem Olajuwon by mentioning that the big man is the only person in league history who is even close to 6,000 combined steals and blocks. Nobody else even has 5,000. I personally love any argument that props up one of my two favorite basketball players of all time. I also like analytics that take the focus off of points per game, which I think might be one of the most overrated stats in all of sports.

Still, before embracing the validity or importance of a stat, I think we have to hold it up to some scrutiny. For years, the RBI was one of the five most important stats in all of baseball. Now we realize that there are too many outside determinants to consider it relevant to a player's worth.

First, we need to consider the source. Bill Simmons is not an expert or analyst of basketball. Other than the fact that he has more access to players, coaches, and executives, his credentials as a basketball expert aren't much better than yours or mine. This isn't say that the stat is irrelevant simply because Bill Simmons is, but we need to be careful about embracing a concept just because some guy managed to get it into print.

Second, most people can agree that an accumulation of stats is somewhat a by-product of remaining in the game for so long. Not to take anything away from those who are able to continue playing at a professional level long beyond what mortal men are capable of (God bless them), but most would agree that gaining 1,800 yards rushing during the prime of a running back's career is a lot more impressive than gaining that same total over the last three or four years of a career, even though they count for the same when considering career totals. In basketball, I prefer the new "per 36 minutes" stats that shows what a player does with his actual time on the court. Better yet, the "per 100 possessions" analytic gives an even more accurate view (in my opinion) because it does not take the pace of the game into effect. Ten years ago, we would have looked at this year's Golden State team and said they have a great offense but their defense is horrible. But once we are able to break things down by 100 possessions, we see that they gave up a lot of points because of the breakneck pace they had going on offense. They were actually giving up a lot fewer points per possession than almost any other team in the league. That's much more meaningful information than just some accumulated total of points allowed.

But do these new stats actually punish players who extend their careers late into their lives? I could be wrong, but I don't think so. Jim Brown and Barry Sanders are the exceptions to the rule. Almost all players wish to extend their professional playing careers as long as they can, whether they are still playing or even capable of playing at an elite level or not. Player A struggles to make a squad and is mostly a reserve for five years. Player B has a short peak where he makes the all-star team once in twelve years and spends half of those years as a quality bench-player. Player C is a future Hall-of-Famer, even though he extended his career for seven years by contributing off the bench. Do we really think there is any way to manipulate the stats to make the careers of Player A or B look better than Player C, even though Player C spent more time as a back-up player than any of them? Sure, there are always outliers, but I think the likelihood is pretty low.

"Stocks" is a big-man's stat. Not every player can be properly evaluated by it. Yes, guards get plenty of steals and can get some blocks, but they are unlikely to even be put in a situation to accumulate the blocks stats that big guys can. Guards certainly have more opportunities to get steals than big men do, but not enough to compensate for their lack of block opportunities. Magic Johnson had a HUGE height advantage over almost every guard he faced, but he still only averaged less than 30 blocks per season. Anfernee Hardaway possessed a similar advantage and only averaged about the same.

The first reason that I even started questioning how important this stat might be is because of Tree Rollins. Tree was another favorite of mine from my youth. He played good defense (even though he was only twice selected to an all-defensive team) and blocked a bunch of shots (even though he led the league in blocks only once). I thought this guy might have some pretty good "stock" numbers (even though he never made an all-star game).

Let's get the obvious out of the way. Olajuwon amassed 2,162 steals and 3,830 blocks, for a total of 5, 992 "stocks." Rollins, who was a backup center for twelve of his eighteen years in the league, only had 512 steals and 2,542 blocks.

Even "per game" this statistic is going to be pretty lopsided seeing Olajuwon was almost always a starter and Rollins played most of his career off the bench. Hakeem's average is 4.8 and Tree's is 2.6. Still, once we take out the accumulator factor, the ratio starts to narrow.

Olajuwon's average per 36 minutes on the court was 4.9, not far from his per game average because Hakeem played close to the league average in minutes for a starter. Tree's average for 36 minutes played was 4.6. Hakeem averaged 6.7 "stocks" per 100 possessions compared to Tree's average of 6.2.

I have to question how a lightly-regarded center who was a back-up for two-thirds of his career compares so well to one of the greatest centers of all time if this stat is so important. It's possible that I just happened to stumble upon the exception to the rule, so I'll see if I can come up with any other decent examples later.

Another good test of the validity of a stat is sometimes to see if other great players have an excellence in that stat in common. As Simmons points out though, even the best of the rest (Robinson, Ewing, Kareem, Mutumbo, and Jordan) were only within 75% of Olajuwon's mark. Of that group, two are considered among the best of all time and the rest are in the Hall-of-Fame (even though only Robinson deserves it).

This is somewhat of a mixed bag but because of the other names that are among the leaders, I would say it helps to legitimize it.

Dikembe's is an interesting scenario also. Here is a guy who won the Defensive Player of the Year award more times than anyone else and was granted entry into the Hall of Fame strictly on the basis of his defensive reputation but his "stocks" rating is lower per game than Olajuwon's and is lower than both Olajuwon and Tree in per 36 minutes and per 100 possessions. There seem to be only two possible conclusions to reach from this. Either this stat has less meaning than Simmons suggests or Dikembe's reputation is considerably overrated. For the record, I'm more willing to believe the latter. Give yourself a clever nickname (like Gary Payton's "The Glove") or a nifty little signature move like the finger-waggle, and it is easy to convince fans and the media that you are better than you really are.

When Simmons is bringing so much light to Olajuwon's "stocks" numbers, he fails to acknowledge that blocks and steals weren't even recorded in the days of Chamberlain and Russell. Certainly Wilt blocked a ridiculous number of shots, and Russell was known to be a pretty craft defender. Who knows if Hakeem has company in that stratosphere of his or not.

Another aspect to consider when deciding whether a stat has much relevance is to see if it has much impact on the outcome of matchups or games. The "defensive rating" tells how many times an opposing player scored on a man in 100 possessions. Anything below 100 is considered to be a pretty good rating. Seeing steals and blocks are both defensive stats (and have a direct impact on possessions), it would make sense that having a lot of "stocks" should correlate to a low defensive rating.

Hakeem led the league in defensive rating five straight seasons (from '86 to '91). During those five seasons his average steals per 100 possessions were either average for his career (2.4) or above (a career high of 3.3). His career average for blocks was 4.3, and in that stretch he had one season of averaging 3.6 blocks, but the rest of the years were above average (up to 4.. His average for "stocks" during this stretch was 7.7, which is definitely above his career average of 6.7, so there definitely seems to be a correlation. Hakeem also ranks in the top 15 for career defense, which also seems to indicate correlation.

Some modern big men that rank above Olajuwon in defense are David Robinson, Ben Wallace, and Tim Duncan, so let's see how they compare to Olajuwon. If they are in the neighborhood or above for "stocks," that would be another indicator in favor of the value of this stat.

Robinson
/36: 4.6
/100: 6.5
Defense 95.6

Wallace
/36: 3.9
/100: 5.8
Defense 95.7

Duncan
/36: 3.1
/100: 4.5
Defense 95.5

Interestingly Robinson's "stock" numbers are quite similar to those of Olajuwon. Duncan and Wallace both fall considerably short. This doesn't necessarily go towards invalidating the stat though because we know there are other ways to play good defense than creating turnovers.

The next (and probably last) criteria we will look at is how well a poor defensive center does in this stat. If it is possible to perform well in "stocks" and still play poor defense, it could go a long ways toward invalidating its worth.

Andrea Bargnani gets a lot of heat for being a terrible defender, and his 111 defensive rating seems to bear that out.

His "stocks" per 36 are 1.6 and per 100 are 3.8.

Even compared to a low "stocks" guy like Tim Duncan, these numbers are pretty low.

I think we are still a ways away from drawing any satisfying conclusions, but my initial research seems to indicate that "stocks" can reflect one type of good defense. A good defender is not always going to be a big "stocks" guy, but a big "stocks" guy will probably be a good defender.