Ken Pomeroy has an interesting piece on his blog examining the question, "How many first-round NBA picks will Kentucky have?" This is a question that has cropped up quite often in recent months due to the largely unprecedented success Kentucky had on the recruiting trail last season.
Kentucky fans will be tempted to "poo-poo" this if it runs afoul of their expectations, which are justifiably high. Kentucky fans have quite rationally adopted John Calipari's pride in seeing relatively large number of first-round picks made from Kentucky's classes each year, and to most fans, it is a mark of success that so many Wildcats get to realize their NBA dreams each year. Senator Mitch McConnel's formulation that Calipari was "... creating more millionaires than a Wall Street firm" is not only a source of pride for Coach Cal, but for most of the Big Blue Nation as well.
What Pomeroy has done here is go back to previous pre-season mock drafts and where the pre-season mock drafts actually wound up come NBA Draft night for the last six NBA drafts. He then performed a regression analysis on the data, and essentially came up with this:
Six years isn’t as much data as I would like, especially since the some of the forecasts from last year can’t be judged yet, but it’s good enough to get us a decent ballpark estimate for each projected pick. According to this analysis, a player projected to be taken 23rd has a 48.9 percent chance of eventually going in the first round, whether it’s the year of the mock draft or some later year.
What this tells you is pretty much what you'd expect — the lower a player is projected in the preseason, the lower his chances of actually getting drafted in the first round on draft day.
What his analysis shows is that, while 7 first-round draft picks from Kentucky is certainly possible, it isn't likely — to the tune of 3.3%. The high probability is that between 4 and 5 Kentucky players will make it into the first round for this season's NBA Draft.
Keep in mind that, like all statistics, what we have here is an attempt to establish a a probability that something will occur. Just because something is not statistically likely doesn't mean it won't happen — we see that every year in college basketball when some school given a low probability of winning a particular game pulls it off anyway, "against the odds." Pomeroy is trying to establish odds of a particular outcome.
He uses the same paradigm for the NBA Draft Lottery to figure out the odds of a given number of lottery picks from Kentucky. The high probability for that question is two, although 3 is comparatively likely enough to be a fair bet.
I think, though, that this goes a bit too far:
Two still turns out to be the best prediction, although three is more likely than one. Anyway, the point here is that saying Kentucky returns two first-rounders and brings in the best freshman class ever is kind of misleading, at least on the former point. And whether this class can match or beat the three lottery picks produced by the Fab Five is far from a guarantee. (Though that’s not the only way to compare the classes.)
I would argue that given the lack of the model's robustness, mainly due to a small sample size, you could rationally argue that three is about as likely as two. Also, I don't really think either statement in his second sentence is misleading in any way, and the comment that three lottery picks is "far from a guarantee" is too clever by half. That same thing can be said about two, or four It's just that two is the most likely outcome.
Also, it is quite likely both Alex Poythress and Willie Cauley-Stein were first-rounders in last year's weak draft, but Poythress, at least, might not be a first-rounder in this season's deeper draft. That kind of variation is not really compensated for in his model, and there's really no good way I see to do that. Some things just can't be modeled.
I agree wholeheartedly with most of what this analysis projects. I do think it is very unlikely that Kentucky can put seven players in the first round, although I do think five is pretty likely, and it turns out to be almost as likely as 4. Six is still a bettable probability if you don't wager much money, in my opinion, but seven is a genuine long-shot.
As for the lottery, I think it is very likely, contra Pomeroy's analysis, that Kentucky puts three players in there. That's not to say that Pomeroy is wrong, the numbers say what they say. But my head (or perhaps it's really my heart and I'm mistaking it for my head) says otherwise, statistics notwithstanding.