Last week in this space, I wrote about inconsistency from a theoretical, logical viewpoint. This week, it's down to the hard facts. I've crunched some numbers in an attempt to answer a pair of questions:

1. Have the Sonics really been that inconsistent this season?
2. Which players' performance has really suffered during back-to-backs?

As you probably know if you're reading this column, the Sonics' blowout loss last Wednesday in Los Angeles to the Clippers dropped them to 0-8 on the season in the second game of back-to-back games. That's a rather remarkable feat. Consider for a second that in all other games, the Sonics have an impressive enough record of 11-7. The odds of a team whose true winning percentage is .611 (that's what 11-7 corresponds with) losing eight consecutive games are about 1 in 2000. So, obviously, there's some other factor at work there. Whether looking at individual output will help us understand what that something else is will be answered soon enough.

For now, however, we turn to the first question. Obviously, it seems like the Sonics have been remarkably inconsistent. Any fan's list of Sonic grievances for Festivus this winter would surely have inconsistency, if not at the top, then somewhere near it. However, whether the numbers bear this out or not is another matter. To measure inconsistency, I'm going to separate a team's points scored and their points allowed. For each, we can find the standard deviation -- a measure of how much each number varies from the average. To take this one step further, I'll divide the standard deviation by the points for or against. Why undertake this additional step? Well, think of it this way. Say we were comparing the deviation in points scored by Art Long to points scored by the Sonics as a team. If somehow we came up that both varied by an average of about two points, which would really be more inconsistent? Obviously, Long's points would be -- two points is a significant portion of the number he scores, whereas it's rather insignificant compared with the team's 90 plus points. In this case, the difference is not so significant, but it's a useful step anyways.

To provide a comparison for the Sonics and give me a base with which to determine how inconsistent they are, I also did the same thing for the other six teams in the Pacific Division. Here are the results, ranked by percentage inconsistency:

LA Clippers - 11.3% Sacramento - 11.3% Golden State - 10.8% Seattle - 10.6% Portland - 10.5% Phoenix - 9% LA Lakers - 8.6%

Surprised by the results? I know I sure was. Looking at their point totals, the Sonics are about average in terms of inconsistency. I'll get to my theory for why we tend to disagree with this in a second. Another shocker was how inconsistent the Kings are. Generally, we tend to think of inconsistency as a bad thing. From a theoretical standpoint, it definitely is for a basketball team as good as the Kings are. Yet they continue sailing along at a merry 18-7. That's one I won't even bother attempting to explain.

Another interesting thing I noticed while looking at the individual statistics for offense and defense was that, in general, there tended to be more inconsistency in points allowed than points scored. This might strike you as a bit odd at first, as it did me -- shouldn't there be an equal amount of variability for each? Well, yes, if we were looking at the NBA as a whole, but breaking it down by team means that this isn't necessarily so. This does have an interesting implication, however. It would seem from this fact that the offense tends to dictate the result of a play more than the defense does. That is, while a team has approximately the same offensive and defensive quality from game to game, the offense translates better to consistent results, making it seem that it is the dominant factor. The Sonics are a bit different in this regard, having about equal numbers for points scored and points allowed. What does this mean? Well, it would seem to say that what inconsistency the Sonics have displayed is due more to going up and down on offense than on defense.

Now, let's return to the most important question of this discussion -- why do we think the Sonics are inconsistent when the numbers don't bear this out? In this case, I think it's because the human brain is comprehending a variable that would be extremely difficult to factor into a computer's logic -- opponent quality. If we were to take off the teams associated with the Sonics' scores, they might not look so inconsistent. As fans, however, we know that the Sonics should not be beating the Lakers on the road one night and then being embarrassed by the Clippers the next. And that, in my opinion, is the source of concern about inconsistency.

To answer the second question, I'm afraid I have a rather inelegant solution. The obvious study to do is to take an average of a player's performance on the first and second nights of back-to-backs, and compare the two. Ideally, this would be done with a comprehensive performance measure, say, for example, the efficiency rating I've devised, which can be found in the stat page. There are two main problems with this, however. Firstly, I don't have the formula with me -- it's on my roommate's computer in my dorm. Secondly, it would take far too long to individually enter the results in each major statistical category from all 16 of the games that have been either the front or back end of a back-to-back this season.

Instead, I used the rather simple category of scoring as a proxy to establish performance level. One possible problem may be obvious -- scoring is correlated to minutes played, whereas an efficiency rating is held independent. My hope is that the fact that the games are taken in pairs of two will tend to neutralize any changes in the rotation, but I know this won't always be true. Another problem is the limited sample size, at best eight for each side. With the Sonics' injuries, many of their regulars have played in only five or so of the back-to-backs.

All that said, I still think that looking at thse numbers can give us some sense of which players have been most adversely affected by back-to-backs. I tracked the Sonics' top eight players thus far, and here is the chart:

After seeing the numbers, some further caveats. Calvin Booth in particular is given unfair treatment. He is one of those who has played in only five sets of back-to-backs. As well, one of the first night games was his 24-point offensive explosion against Dallas that, so far as anyone can tell, is the exception, not the rule. Take that game out, and he doesn't seem so horribly adversely affected.

There appear to be two players in particular who are demonstrating the negative effects of playing too many minutes through their scoring numbers -- Rashard Lewis and Gary Payton. Indeed, Payton leads the Sonics in minutes played, while Lewis is third. Payton's always played a lot of minutes, but he hasn't always been 33, and I don't think he's capable of bouncing back as quickly anymore. That said, it is true that Payton's numbers are probably slightly affected by the fact that the Sonics have so often been blown out in the second half of back-to-backs.

Lewis, like Booth, had a fluke explosion on the first night -- his career-high 36 points against Orlando on the East Coast swing -- but even this is far from enough to explain the downturn he's seen. Not only have Lewis' minutes increased this season, but he's also been asked to spend far more time as the power forward, banging with much heavier and stronger players. Fatigue is inevitable under these circumstances.

The remedy for this problem is quite simple -- cutting the minutes for both Lewis and Payton, especially on the first night of back-to-backs -- but implementation is more difficult. The fact is that in all but one of those first games, the Sonics have been in close games down to the final minutes. Certainly, Coach McMillan wants to do everything he can to get that first game, which the Sonics have successfully done Five out of Eight times. Twice, even more devestating, the first game has gone to overtime.

If Lewis and Payton must be played the first night, then the only option is to lean more heavily on the bench in the second. Rookies Earl Watson and Vladimir Radmanovic have both proven themselves more then capable in recent games (though now Radmanovic is hurting). If ineffectiveness is noticed in the starters, a quicker trigger to the bench is necessary, to keep the deficit from getting so ridiculous that the energy provided by this duo, Desmond Mason, and Art Long, is not too little, too late.