Sunday, August 16, 2015

Tiger 2000 >>> Jordan 2015

One of the announcers on the early coverage of the last round of the PGA Championship suggested that if Jordan Spieth were to win, for his third major championship of the year, he would have had the best year ever. Ahem. This is ridiculous. Here's what Spieth has done so far this year. He's won four tournaments, two of them majors, the others both relatively minor tournaments (the Valspar Championship and the John Deere Classic). He's finished second three times, third once, and in the top 10 thirteen times overall out of 20 tournaments. He's missed two cuts, one of them at the Players Championship, and has finished outside the top 25 four times in all. Now, that's a really, really great season.

But in 2000, Tiger Woods won, including three consecutive majors, nine total tournaments. He won his three majors by a combined 23 shots, compared to the 5 by which Spieth won his first two (and he doesn't look to add more than another couple today, if he does win). Overall Tiger won those nine tournaments by a combined margin of 46 shots, including three wins by at least 8 shots; both of Spieth's non-major wins were in playoffs, so he's got a combined margin of victory of just 5. Also, Tiger missed zero cuts; in fact, he didn't once finish outside the top 25! Only three times did he miss the top 10, compared to Spieth's seven (and counting). He led the Tour in GIR%, holes per eagle, birdies per round, scoring average (adjusted or otherwise), total driving, ball-striking, all-around ranking, par-3 scoring average, par-4 scoring average, par-5 scoring average, birdie percentage on par 4s and on par 5s, bounce-back, birdie conversion percentage, par breaker percentage, scoring average in rounds 1, 2, and 3, front and back nine scoring average, total eagles, lowest round, and, of course, the official money list. He was also World #1, by an enormous margin. Like, practically a 3-1 margin. Also he finished second in driving distance, putts per GIR, and final round scoring average, and was third in scrambling. His all-around ranking (the sum of his ranks in eight key statistics) was 40% lower than the second-place finisher, David Duval.

Spieth, meanwhile, leads the Tour in the following statistics, of those that existed in 2000: birdie average, scoring average (actual and adjusted), putts per GIR, one-putt percentage, putts per round, putts per round in round 2, lowest round, par breaker percentage, round 2 scoring average, back 9 scoring average, scoring average and birdie percentage on par 4s, lowest average, and, of course, the money list. That's a lot! But it's not as much as Tiger; in fact, Tiger led every single one of those stats except for the putting stats (and he was second in putts per GIR, the best putting stat that existed back in 2000). Meanwhile, Spieth is 55th in GIR%, 42nd in total driving (including 76th in distance and 85th in accuracy), 25th in par-3 scoring, 20th in par-5 scoring. Oh, and he's third in all-around ranking, with a figure about 30% above Will Wilcox's leading figure. (What?!)

The point is that, obviously, Jordan Spieth is a great player, and is having a great year, but he isn't coming close to dominating the Tour, either in terms of performance or in terms of statistics, the way Tiger did. If we're talking best years ever, and we're excluding Bobby Jones's 1930 for some reason and don't consider Hogan's 1953, or any of the years when he or Byron Nelson or whoever won double digits of tournaments, then there just isn't any question. Tiger's 2000 was the best year anyone has ever had, and I daresay it will continue to be for a long, long time.

Thursday, August 13, 2015

Update on the Bartolo Colon Saberhagenometer

It's, uh, not going well. In our last update, which was also the first update, Bartolo Colon had five wins on the season against just one walk. Since then, though, he's got a 5-10 record and has walked 13 batters, dropping him to 10 wins and 14 walks on the season. His Saberhagen score is, thus, -4. Now, he's closed that gap a little bit recently, with no walks in his last four starts, one of which he won. And if he were to pitch really really well the rest of the way it wouldn't be crazy to see him climb back toward positive territory here. But it's really not very likely. If you want a Saberhagen, you really can't afford three-walk games, or back-to-back multi-walk games, or five starts in a row with a walk. Bartolo's had all three (and none of them overlapped). So, it's not looking good.

However! Five of his walks have been of the intentional variety, and in a sense it's not really fair to penalize him for them. In a sense it is, as it was his pitching that created the need to issue the walk, but they also don't represent his inability to throw strikes or anything like that. So if we define the Modified Saberhagen, defined as Wins minus (Walks minus Intentional Walks), then Colon is still at +1, and has a much more reasonable shot at it. Interestingly, Saberhagen had no intentional walks in 1994, so his Modified Saberhagen score was identical to his normal Saberhagen score at +1. Of course, the Modified Saberhagen isn't really as impressive as the proper Saberhagen, but it's still got some validity.

It's a bit trickier to get a sense of who else has pulled off the Modified Saberhagen, since the "unintentional walks" stat often isn't kept as its own thing. I was able to make a rough stab at it using some custom FanGraphs leaderboards, though, which reveal two other completed seasons that just barely pulled it off: 2005 Carlos Silva (9 W, 9 BB, 2 IBB) and 2014 Phil Hughes (16 W, 16 BB, 1 IBB). Then there's 1997 Greg Maddux, who won 19 games and walked 20 batters, but six of those were intentional, giving him a Modified Saberhagen of +5! As much as all the other Saberhageners put together, counting 2015 Bartolo Colon. Maddux finished second in the Cy Young Award voting that year, to Pedro Martinez's last season with the Expos, and deservedly so, but only because peak Pedro was insane.

Now, this list only includes players who qualified for the ERA title and who played at a time when they keep track of intentional walks. So there may be others who have pulled off the Modified Saberhagen who don't show up here. (I could check for those with fewer innings, but that would just be more work so I'm not going to.) I think it's pretty clear that the Modified Saberhagen is orders of magnitude less impressive than the proper one, even if Saberhagen himself didn't care about the difference. Still, at least Colon's season isn't entirely lost.

The Connecticut Supreme Court is Right: Prospective Abolition of the Death Penalty is Cruel and Unusual

In 2012, the Connecticut legislature abolished the death penalty. Today, the Connecticut Supreme Court held that the death penalty violated the state constitution. You may be wondering how both of those sentences can be true (and no, they didn't reinstate it during the past three years). The answer is that the 2012 abolition was prospective; that is, it did not apply to those who had already been convicted and sentenced to death. The court's ruling, then, was that actually carrying out the death sentences that have already been imposed on Connecticut's 11 death row inmates would be (state-)unconstitutional. More specifically they held that executions would violate the due process provisions of the Connecticut constitution, which are understood to contain a ban on cruel and unusual punishments. Now, there's nothing especially remarkable about the idea that capital punishment is cruel and unusual. But the court's reasoning was, at least in part, new and interesting, and I quite like the move they make.

The basic puzzle of what, at the federal level, we call Eighth Amendment jurisprudence is determining what the second part of "cruel and unusual" means. Well, that's the puzzle if we reject the Scalia-esque "originalist"* notion that the phrase "cruel and unusual punishments" is just a shorthand for a discrete list of punishments considered cruel and unusual in 1791. Which we do. So the question then becomes, okay, we have a sense of what a "cruel" punishment is, but how to measure its unusuality? Do we just count the states? By raw number of jurisdictions, or by population? Do foreign countries count, and if so which ones? How quickly do new trends away from the use of a certain punishment become incorporated into the Constitution? Are public opinion polls relevant? All of this arises primarily in the federal context, wherein the U.S. Supreme Court is interpreting the Eighth Amendment as it applies to the federal government directly and to the states as incorporated into the Fourteenth Amendment. But analogous issues could arise when a state court interprets an analogous provision of the state constitution. How do the other 49 states get counted for that purpose? Do they play the same role as in federal Eighth Amendment analysis, or are they more like foreign nations? After all, they are external to the individual state. But within a given state a punishment won't be "common" or "unusual," it will either be provided for by state law or it won't be; internal counting becomes binary. The whole thing is rather a mess.**


The neat thing about this Connecticut case, though, is that it doesn't have to bother with that whole issue. Because what they say, at least in part, is that we know executing these condemned prisoners would be cruel and unusual because the Connecticut legislature has already said so. They abolished the death penalty! Because they thought it was cruel! Making it, thereafter, not to be found within the state of Connecticut. At all. Which is rather on the unusual side. This is brilliant, and, I think, surely right. Is it not truly anomalous, truly unusual, for a state to inflict a punishment which it has already abolished? Is there not something singularly cruel about saying to a handful of unlucky people, we now think it's wrong to punish anyone as we have determined to punish you, but we will so punish you nonetheless? It's, like, sick.*** It almost has a somewhat bill-of-attainder feel to it, i.e., "we will not impose capital punishment (except on these eleven specific individuals)."

One interesting thing, though, is the dynamic if the rule against prospective abolition (or rather, the rule that any purported prospective abolition must in fact be retroactive as well) became firmly established. Because presumably this would discourage any state that was considering a prospective abolition from doing so--assuming, of course, that there really wouldn't be the appetite for retroactive abolition in that state. So while in principle I might like it if the U.S. Supreme Court held that a state which has prospectively abolished the death penalty cannot carry out executions going forward (again, assuming they're not gonna just do the right thing and hold the whole thing unconstitutional), there might be some interesting prudential reasons not to do that, but rather to hope that the courts of each prospective-abolitionist state will surprise their legislatures with a ruling like this one out of Connecticut. Except that many state courts might come out the other way, and if not then at some point legislatures might wise up anyway. There's sort of an interesting game theory dynamic to the whole thing.

But the basic point is that this is a very clever and well-reasoned decision, and of course a humane and decent one. Also it makes me happy because it was a state law decision and therefore cannot be overruled by the federal courts. Hooray for state constitutional law!


*Asterisk meant to note that Scalia is by no means the exclusive and authoritative oracle of originalism, and other originalists, like Steven Calabresi, might not endorse his approach.
**Of course, this is one of the reasons why I like Justice William Brennan's approach from Furman v. Georgia, which understands "unusual" not to mean "rare" or "infrequent" but to mean "strange" or "weird." Thus he held capital punishment to be cruel and unusual because we long ago abandoned all other forms of corporal punishment, making execution qualitatively unlike any other punishment currently in use within America. Of course, this qualitative determination becomes a matter for the reasoned judgment of judges, rather than imposing any external/objective constraints on judges, but hopefully the paragraph above demonstrates that the quantitative approach is so ill-defined that it doesn't really constrain judges either. Also, y'know, the idea that we need to be constraining judges is itself wrong-headed, but that's sort of a broader point.
***Of course, this isn't the only context in which laws are passed reducing the available punishments for certain crimes on a prospective basis. Recently, for instance, both the federal government and many states have reduced sentences for various drug crimes, but have not automatically released every drug criminal who had already served more than the new maximum sentence for their crime. This feels less appalling, though, and I think the reason why must be found in Brennan's qualitative approach. Connecticut divested itself of the power to impose an entire class of punishment, except for a handful of unlucky souls. By contrast, reducing prison sentences for drug crimes does not strip the state of its power to imprison people. There is also the singular quality of executions, that they are carried out at a specific time subsequent to their imposition. Indeed actually carrying out an execution is commonly understood to be a distinct decision, separate from the decision to impose the sentence; hence, there is something singularly ill-seeming about making the decision to carry out an execution when the state can no longer make the decision to impose such a sentence. I would imagine, however, that other prospective abolitions of whole categories of punishment would raise similar issues.